DL-Art-School/codes/models/archs/arch_util.py

391 lines
14 KiB
Python
Raw Normal View History

2019-08-23 13:42:47 +00:00
import torch
import torch.nn as nn
import torch.nn.init as init
import torch.nn.functional as F
import torch.nn.utils.spectral_norm as SpectralNorm
from math import sqrt
2019-08-23 13:42:47 +00:00
def pixel_norm(x, epsilon=1e-8):
return x * torch.rsqrt(torch.mean(torch.pow(x, 2), dim=1, keepdims=True) + epsilon)
2019-08-23 13:42:47 +00:00
def initialize_weights(net_l, scale=1):
if not isinstance(net_l, list):
net_l = [net_l]
for net in net_l:
for m in net.modules():
if isinstance(m, nn.Conv2d) or isinstance(m, nn.Conv3d):
2019-08-23 13:42:47 +00:00
init.kaiming_normal_(m.weight, a=0, mode='fan_in')
m.weight.data *= scale # for residual block
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, nn.Linear):
init.kaiming_normal_(m.weight, a=0, mode='fan_in')
m.weight.data *= scale
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, nn.BatchNorm2d):
init.constant_(m.weight, 1)
init.constant_(m.bias.data, 0.0)
2020-06-06 03:02:08 +00:00
def make_layer(block, n_layers, return_layers=False):
2019-08-23 13:42:47 +00:00
layers = []
for _ in range(n_layers):
layers.append(block())
2020-06-06 03:02:08 +00:00
if return_layers:
return nn.Sequential(*layers), layers
else:
return nn.Sequential(*layers)
2019-08-23 13:42:47 +00:00
class ResidualBlock(nn.Module):
'''Residual block with BN
---Conv-BN-ReLU-Conv-+-
|________________|
'''
def __init__(self, nf=64):
super(ResidualBlock, self).__init__()
self.lrelu = nn.LeakyReLU(negative_slope=0.1, inplace=True)
self.conv1 = nn.Conv2d(nf, nf, 3, 1, 1, bias=True)
self.BN1 = nn.BatchNorm2d(nf)
self.conv2 = nn.Conv2d(nf, nf, 3, 1, 1, bias=True)
self.BN2 = nn.BatchNorm2d(nf)
# initialization
initialize_weights([self.conv1, self.conv2], 0.1)
def forward(self, x):
identity = x
out = self.lrelu(self.BN1(self.conv1(x)))
out = self.BN2(self.conv2(out))
return identity + out
class ResidualBlockSpectralNorm(nn.Module):
'''Residual block with Spectral Normalization.
---SpecConv-ReLU-SpecConv-+-
|________________|
'''
def __init__(self, nf, total_residual_blocks):
super(ResidualBlockSpectralNorm, self).__init__()
self.lrelu = nn.LeakyReLU(negative_slope=0.1, inplace=True)
self.conv1 = SpectralNorm(nn.Conv2d(nf, nf, 3, 1, 1, bias=True))
self.conv2 = SpectralNorm(nn.Conv2d(nf, nf, 3, 1, 1, bias=True))
initialize_weights([self.conv1, self.conv2], 1)
def forward(self, x):
identity = x
out = self.lrelu(self.conv1(x))
out = self.conv2(out)
return identity + out
2019-08-23 13:42:47 +00:00
class ResidualBlock_noBN(nn.Module):
'''Residual block w/o BN
---Conv-ReLU-Conv-+-
|________________|
'''
def __init__(self, nf=64):
super(ResidualBlock_noBN, self).__init__()
self.lrelu = nn.LeakyReLU(negative_slope=0.1, inplace=True)
2019-08-23 13:42:47 +00:00
self.conv1 = nn.Conv2d(nf, nf, 3, 1, 1, bias=True)
self.conv2 = nn.Conv2d(nf, nf, 3, 1, 1, bias=True)
# initialization
initialize_weights([self.conv1, self.conv2], 0.1)
def forward(self, x):
identity = x
out = self.lrelu(self.conv1(x))
2019-08-23 13:42:47 +00:00
out = self.conv2(out)
return identity + out
def flow_warp(x, flow, interp_mode='bilinear', padding_mode='zeros'):
"""Warp an image or feature map with optical flow
Args:
x (Tensor): size (N, C, H, W)
flow (Tensor): size (N, H, W, 2), normal value
interp_mode (str): 'nearest' or 'bilinear'
padding_mode (str): 'zeros' or 'border' or 'reflection'
Returns:
Tensor: warped image or feature map
"""
assert x.size()[-2:] == flow.size()[1:3]
B, C, H, W = x.size()
# mesh grid
grid_y, grid_x = torch.meshgrid(torch.arange(0, H), torch.arange(0, W))
grid = torch.stack((grid_x, grid_y), 2).float() # W(x), H(y), 2
grid.requires_grad = False
grid = grid.type_as(x)
vgrid = grid + flow
# scale grid to [-1,1]
vgrid_x = 2.0 * vgrid[:, :, :, 0] / max(W - 1, 1) - 1.0
vgrid_y = 2.0 * vgrid[:, :, :, 1] / max(H - 1, 1) - 1.0
vgrid_scaled = torch.stack((vgrid_x, vgrid_y), dim=3)
output = F.grid_sample(x, vgrid_scaled, mode=interp_mode, padding_mode=padding_mode)
return output
class PixelUnshuffle(nn.Module):
def __init__(self, reduction_factor):
super(PixelUnshuffle, self).__init__()
self.r = reduction_factor
def forward(self, x):
(b, f, w, h) = x.shape
x = x.contiguous().view(b, f, w // self.r, self.r, h // self.r, self.r)
x = x.permute(0, 1, 3, 5, 2, 4).contiguous().view(b, f * (self.r ** 2), w // self.r, h // self.r)
return x
2020-07-03 18:06:38 +00:00
# simply define a silu function
def silu(input):
'''
Applies the Sigmoid Linear Unit (SiLU) function element-wise:
SiLU(x) = x * sigmoid(x)
'''
2020-07-05 23:28:00 +00:00
return input * torch.sigmoid(input)
# create a class wrapper from PyTorch nn.Module, so
# the function now can be easily used in models
class SiLU(nn.Module):
'''
Applies the Sigmoid Linear Unit (SiLU) function element-wise:
SiLU(x) = x * sigmoid(x)
Shape:
- Input: (N, *) where * means, any number of additional
dimensions
- Output: (N, *), same shape as the input
References:
- Related paper:
https://arxiv.org/pdf/1606.08415.pdf
Examples:
>>> m = silu()
>>> input = torch.randn(2)
>>> output = m(input)
'''
def __init__(self):
'''
Init method.
'''
super().__init__() # init the base class
def forward(self, input):
'''
Forward pass of the function.
'''
2020-07-05 23:28:00 +00:00
return silu(input)
2020-07-03 18:06:38 +00:00
''' Convenience class with Conv->BN->ReLU. Includes weight initialization and auto-padding for standard
kernel sizes. '''
class ConvBnRelu(nn.Module):
def __init__(self, filters_in, filters_out, kernel_size=3, stride=1, activation=True, norm=True, bias=True):
2020-07-03 18:06:38 +00:00
super(ConvBnRelu, self).__init__()
padding_map = {1: 0, 3: 1, 5: 2, 7: 3}
assert kernel_size in padding_map.keys()
self.conv = nn.Conv2d(filters_in, filters_out, kernel_size, stride, padding_map[kernel_size], bias=bias)
if norm:
2020-07-03 18:06:38 +00:00
self.bn = nn.BatchNorm2d(filters_out)
else:
self.bn = None
if activation:
2020-07-03 18:06:38 +00:00
self.relu = nn.ReLU()
else:
self.relu = None
# Init params.
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu' if self.relu else 'linear')
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.conv(x)
if self.bn:
x = self.bn(x)
if self.relu:
return self.relu(x)
else:
return x
''' Convenience class with Conv->BN->SiLU. Includes weight initialization and auto-padding for standard
kernel sizes. '''
class ConvBnSilu(nn.Module):
def __init__(self, filters_in, filters_out, kernel_size=3, stride=1, activation=True, norm=True, bias=True, weight_init_factor=1):
super(ConvBnSilu, self).__init__()
padding_map = {1: 0, 3: 1, 5: 2, 7: 3}
assert kernel_size in padding_map.keys()
self.conv = nn.Conv2d(filters_in, filters_out, kernel_size, stride, padding_map[kernel_size], bias=bias)
if norm:
self.bn = nn.BatchNorm2d(filters_out)
else:
self.bn = None
if activation:
self.silu = SiLU()
else:
self.silu = None
# Init params.
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu' if self.silu else 'linear')
m.weight.data *= weight_init_factor
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.conv(x)
if self.bn:
x = self.bn(x)
if self.silu:
return self.silu(x)
else:
return x
2020-07-03 18:06:38 +00:00
''' Convenience class with Conv->BN->LeakyReLU. Includes weight initialization and auto-padding for standard
kernel sizes. '''
class ConvBnLelu(nn.Module):
def __init__(self, filters_in, filters_out, kernel_size=3, stride=1, activation=True, norm=True, bias=True, weight_init_factor=1):
2020-07-03 18:06:38 +00:00
super(ConvBnLelu, self).__init__()
padding_map = {1: 0, 3: 1, 5: 2, 7: 3}
assert kernel_size in padding_map.keys()
self.conv = nn.Conv2d(filters_in, filters_out, kernel_size, stride, padding_map[kernel_size], bias=bias)
if norm:
2020-07-03 18:06:38 +00:00
self.bn = nn.BatchNorm2d(filters_out)
else:
self.bn = None
if activation:
2020-07-03 18:06:38 +00:00
self.lelu = nn.LeakyReLU(negative_slope=.1)
else:
self.lelu = None
# Init params.
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, a=.1, mode='fan_out',
nonlinearity='leaky_relu' if self.lelu else 'linear')
m.weight.data *= weight_init_factor
if m.bias is not None:
m.bias.data.zero_()
2020-07-03 18:06:38 +00:00
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.conv(x)
if self.bn:
x = self.bn(x)
if self.lelu:
return self.lelu(x)
2020-07-07 02:59:59 +00:00
else:
return x
''' Convenience class with Conv->GroupNorm->LeakyReLU. Includes weight initialization and auto-padding for standard
kernel sizes. '''
class ConvGnLelu(nn.Module):
def __init__(self, filters_in, filters_out, kernel_size=3, stride=1, activation=True, norm=True, bias=True, num_groups=8, weight_init_factor=1):
2020-07-07 02:59:59 +00:00
super(ConvGnLelu, self).__init__()
padding_map = {1: 0, 3: 1, 5: 2, 7: 3}
assert kernel_size in padding_map.keys()
self.conv = nn.Conv2d(filters_in, filters_out, kernel_size, stride, padding_map[kernel_size], bias=bias)
if norm:
2020-07-07 02:59:59 +00:00
self.gn = nn.GroupNorm(num_groups, filters_out)
else:
self.gn = None
if activation:
2020-07-07 02:59:59 +00:00
self.lelu = nn.LeakyReLU(negative_slope=.1)
else:
self.lelu = None
# Init params.
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, a=.1, mode='fan_out',
nonlinearity='leaky_relu' if self.lelu else 'linear')
m.weight.data *= weight_init_factor
if m.bias is not None:
m.bias.data.zero_()
2020-07-07 02:59:59 +00:00
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.conv(x)
if self.gn:
x = self.gn(x)
if self.lelu:
return self.lelu(x)
else:
return x
''' Convenience class with Conv->BN->SiLU. Includes weight initialization and auto-padding for standard
kernel sizes. '''
class ConvGnSilu(nn.Module):
def __init__(self, filters_in, filters_out, kernel_size=3, stride=1, activation=True, norm=True, bias=True, num_groups=8, weight_init_factor=1):
super(ConvGnSilu, self).__init__()
padding_map = {1: 0, 3: 1, 5: 2, 7: 3}
assert kernel_size in padding_map.keys()
self.conv = nn.Conv2d(filters_in, filters_out, kernel_size, stride, padding_map[kernel_size], bias=bias)
if norm:
self.gn = nn.GroupNorm(num_groups, filters_out)
else:
self.gn = None
if activation:
self.silu = SiLU()
else:
self.silu = None
# Init params.
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu' if self.silu else 'linear')
m.weight.data *= weight_init_factor
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
def forward(self, x):
x = self.conv(x)
if self.gn:
x = self.gn(x)
if self.silu:
return self.silu(x)
2020-07-03 18:06:38 +00:00
else:
return x
# Block that upsamples 2x and reduces incoming filters by 2x. It preserves structure by taking a passthrough feed
# along with the feature representation.
class ExpansionBlock(nn.Module):
2020-07-11 04:57:34 +00:00
def __init__(self, filters_in, filters_out=None, block=ConvGnSilu):
super(ExpansionBlock, self).__init__()
2020-07-11 04:57:34 +00:00
if filters_out is None:
filters_out = filters_in // 2
self.decimate = block(filters_in, filters_out, kernel_size=1, bias=False, activation=False, norm=True)
self.process_passthrough = block(filters_out, filters_out, kernel_size=3, bias=True, activation=False, norm=True)
2020-07-11 05:00:21 +00:00
self.conjoin = block(filters_out*2, filters_out, kernel_size=3, bias=False, activation=True, norm=False)
2020-07-11 04:57:34 +00:00
self.process = block(filters_out, filters_out, kernel_size=3, bias=False, activation=True, norm=True)
# input is the feature signal with shape (b, f, w, h)
# passthrough is the structure signal with shape (b, f/2, w*2, h*2)
# output is conjoined upsample with shape (b, f/2, w*2, h*2)
def forward(self, input, passthrough):
x = F.interpolate(input, scale_factor=2, mode="nearest")
x = self.decimate(x)
p = self.process_passthrough(passthrough)
x = self.conjoin(torch.cat([x, p], dim=1))
return self.process(x)